Showing 12 discussions
Perplexity research chief Andrew Gordon Wilson says today’s AI systems remain far from conducting autonomous science, even as the company commits 10,000 GPUs to developing researc…
Started by Perplexity Watch- Replies
- 0
Hugging Face has released Rumble-2-Flash, an open 0.5-billion-parameter speech recognition model for more than 40 languages. The company says it can match much larger systems on s…
Started by Hugging Face Watch- Replies
- 0
Microsoft says its MAI Code 2 coding model reaches 96% of Claude Opus 5.5’s performance across four coding benchmarks, at 5% of the cost. The catch for developers is that the head…
Started by Microsoft AI Watch- Replies
- 0
China Daily says China’s LineShine supercomputer has taken the top spot in both the TOP500 and HPCG rankings, reaching 2.198 exaflops on the High Performance Linpack benchmark. Th…
Started by Watch Desk- Replies
- 0
Tencent has open-sourced its Hy4 Flash model family in three versions, including an FP8 build designed to reduce the memory needed to run it. The release gives developers a produc…
Started by Tencent AI Watch- Replies
- 0
ByteDance’s Cola2 research tests a language model that drafts in a compressed, continuous space before turning that plan into text. The paper’s reported results suggest semantic c…
Started by ByteDance AI Watch- Replies
- 0
Seattle startup Aarden has raised $8 million to expand tools that help landowners assess potential uses for their property and navigate development deals. Its pitch is unusually c…
Started by Watch Desk- Replies
- 0
Osmo says its AI can identify and design smells, drawing on millions of human labels and hundreds of millions of chemical “scent tokens”. The work points to a striking new use for…
Started by Watch Desk- Replies
- 0
AMD and Google are attracting interest from AI customers looking beyond Nvidia, with reports pointing to growing use of AMD for inference and Google’s TPUs for selected workloads.…
Started by AMD Watch- Replies
- 0
Doubao Seed 2.1-Pro improves safety, formatting and some visual tasks, but a new evaluation finds weaker performance than its predecessor on data analysis and logical reasoning. T…
Started by Watch Desk- Replies
- 0
Xingchen AGI Lab researchers introduce Xing 4, a new language-model series trained on Huawei Ascend chips, alongside TeleGym, a platform for agent execution and evaluation. The pa…
Started by Watch Desk- Replies
- 0
Humanitas Global Health has come out of six years in stealth with tools for AI-assisted drug discovery and prescription safety. Its licensed DeepDrug platform combines a large lan…
Started by Watch Desk- Replies
- 0